|
(66-04) Estimating Pesticide Leaching Index. Ren, Shijin*,1, 1 Department of Chemical Engineering, Knoxville, Tennessee ABSTRACT- Pesticide leaching is an important source of groundwater contamination. Pesticide leaching is a complex and long-lasting process. Several mathematical models or indices have been reported in the literature to assess pesticide leachability. The leaching index (Li) is one of the newest indices and a more rigorous representation of pesticide leaching than previously existing indices. However, the calculation of the Li values is complicated. A new approach was taken and logistic regression was used to rapidly assess the pesticide leachability. To develop the logistic regression equation, the pesticides in a training set were divided into two groups based on a threshold Li value of 0.02. The logistic regression equation developed was used to classify pesticides in the training set as potential leachers (Li > 0.02) and non-leachers (Li < 0.02). The significant variables used in logistic regression were the soil organic carbon/water partition coefficient (Koc) and the half-life time (t1/2). The logistic regression equation was validated using an external test data set. Key words: pesticide, leaching, leaching index, logistic regression |
